摘要 |
PURPOSE: A service level agreement management method for guarantee of quality of service in cloud environment is provided to perform services by dividing the services into the other server when excessive service requests are asked. CONSTITUTION: A cloud service user makes a virtual network template with a virtual machine through a web interface and then delivers the template to a front end server for service request (S10). The front end server selects a cluster node (S20), and makes a template file of the virtual machine and a template file of a virtual network, and generates the virtual network and the virtual machine in a corresponding node (S30). When the virtual machine allocated to the node becomes in an operation state, the user is able to use a cloud service and the front end server availability, response time and processing rate of the virtual machine used by the user (S40). If index threshold is deviated from, a message is sent to the front end server so that recovery appropriate for the corresponding message request is performed (S50). If the index threshold is satisfied, the front end server performs real-time monitoring of the virtual machines (S60). [Reference numerals] (AA) Start; (BB) Select a cluster node; (CC,II) Yes; (DD) ACPU >= LCPU of the selected node; (EE,JJ) Yes; (FF) Restore; (GG) Sense faults; (HH) Three indexes (availability, processing rate, and response time) <= Threshold; (S10) Request the generation of a virtual machine of a user; (S30) Generate the virtual machine after making a template file; (S40) Measure the availability, the processing rate, and the response time of the respective virtual machine; (S60) Monitor |